This study will use methods, techniques, and findings from the fields of affect and emotion, behavioral assessment, and psychometrics to shed light on a core feature of BPD, affective instability. In this revised application, we propose to conduct an intensive study of affective instability using ecological momentary assessment (EMA) --- a real time assessment of behaviors, emotions, and cognitive variables via hand held computers. Study participants will include 75 BPD outpatients and 75 psychiatric controls (with Major Depressive Disorder) who will rate their mood states, behaviors, and life events six times per day for a 28 day period. We will address four major aims: (1) characterize the mood state patterns of BPD patients and contrast these patterns with those of patients of near-neighbor diagnoses (in this case major depression); (2) identify antecedents (life events) and consequents (substance use behaviors) of mood shifts; (3) determine whether BPD patients can accurately recall (via retrospective self-report instruments) their mood shifts and variables associated with these shifts; and (4) assess whether existing measures of affective instability, affective intensity, and personality are related to EMA measures of affective variability and mood shifts. This study will represent the most intensive, naturalistic study of affective instability, and it will address several theoretical questions regarding affective instability. Further, these results will have implications for treatment and for the assessment of affective instability. Funding for this study will allow us to provide multidisciplinary training to an advanced graduate student who plans to conduct research in this area in the future. Completion of this study will also enable us to conduct multiple studies aimed at elucidating the utility of EMA in the assessment of BPD and the utility of EMA in tracking treatment effects, as well as to take part in the development of a multidisciplinary center for the study of BPD and its features that will foster collaborations among scientists with expertise in psychometrics and measurement theory, psychobiology, social and affective neuroscience, and basic behavioral science. [unreadable] [unreadable]